Method and apparatus for inversion detection

Abstract

A method for inversion counting or phlebotomist monitoring can include identifying, by processing circuitry, whether a blood collection tube is present in image data from a camera situated to capture images of a phlebotomist collecting a sample, in response to identifying the blood collection tube is present in the field of view of the camera based on the image data, identifying whether the blood collection tube includes blood therein, after identifying the blood is present in the blood collection tube, counting, based on the image data, a number of inversions performed on the blood collection tube, and in response to determining the number of inversions performed is less than a required number of inversions, issuing an alert indicating that insufficient inversions were performed.

I claim: 
     
       1. A system comprising:
 a camera to capture image data; 
 an analysis system comprising processing circuitry and a memory, the memory including (i) data indicating a number of inversions required to be performed on a blood collection tube and (ii) instructions that when executed by the processing circuitry, cause the processing circuitry to perform operations for inversion monitoring, the operations comprising: 
 identifying, using object recognition, whether a blood collection tube is present in the image data; 
 in response to identifying the blood collection tube is present in the field of view of the camera based on the image data, identifying whether the blood collection tube includes blood therein; 
 after identifying the blood is present in the blood collection tube, counting, based on the image data and using a computer model trained to distinguish between an upright state of the blood collection tube and an upside down state of the blood collection tube, a number of inversions performed on the blood collection tube; and 
 in response to determining the number of inversions performed is less than the required number of inversions, issuing, by communications circuitry, an alert indicating that insufficient inversions were performed. 
 
     
     
       2. The system of  claim 1 , wherein identifying whether a blood collection tube is present in the image data includes comparing a feature identified in the image data to features of blood collection tubes in the memory. 
     
     
       3. The system of  claim 1 , wherein identifying whether the blood collection tube includes blood therein includes determining whether a body portion of the blood collection tube has changed color in a series of images. 
     
     
       4. The system of  claim 1 , wherein the operations further comprise identifying a number of inversions required to be performed on the blood collection container. 
     
     
       5. The system of  claim 4 , wherein identifying the number of inversions required includes identifying a color of the cap of the blood collection container and looking up the number of inversions associated with the identified color in the memory. 
     
     
       6. The system of  claim 5 , wherein identifying the number of inversions required includes identifying whether a material is present in a body portion of the blood collection container. 
     
     
       7. The system of  claim 1 , wherein counting the number of inversions performed on the blood collection tube includes identifying a first feature of the blood collection tube on a cap of the blood collection tube and a second feature on an end portion of the blood collection tube opposing the cap, and for each inversion, determining that pixels corresponding to the first feature are vertically displaced from the second feature in a first direction and then vertically displaced from the second feature in a second, opposite direction in image data corresponding to a subsequent image. 
     
     
       8. The system of  claim 1 , wherein counting the number of inversions performed on the blood collection tube includes identifying in a quadrant of a logical grid in which a feature is detected and then identifying the feature in an opposing quadrant in image data of a subsequent image, the logical grid non-rotatable and formed at a center of the blood container. 
     
     
       9. The system of  claim 1 , wherein the model is trained based on images of blood collection tubes labelled upright and upside down and counting the number of inversions performed on the blood collection tube includes detecting in image data of subsequent images the blood collection container is upright and the blood collection container is upside down, respectively. 
     
     
       10. The system of  claim 1 , wherein the operations further comprise identifying and recording, in the memory, information on a label of the blood collection tube including identifying initials of the phlebotomist and recording a phlebotomist identification associated with the phlebotomist along with the counted number of inversions for each blood collection tube in the memory. 
     
     
       11. The system of  claim 1 , wherein the operations further comprise using a facial recognition technique to identify a phlebotomist in the image data and recording a phlebotomist identification associated with the phlebotomist along with the counted number of inversions for each blood collection tube in the memory. 
     
     
       12. The system of  claim 1 , wherein the operations further comprise determining, based on the image data, whether there is a sufficient volume of blood in the blood collection tube to perform a specified test. 
     
     
       13. A non-transitory machine-readable medium including instructions that, when executed by a machine, cause the machine to perform operations for inversion counting and monitoring, the operations comprising:
 identifying, using object recognition, whether a blood collection tube is present in image data from a camera situated to capture a series of images of a phlebotomist collecting a sample; 
 in response to identifying the blood collection tube is present in the field of view of the camera based on the image data, identifying whether the blood collection tube includes blood therein; 
 after identifying the blood is present in the blood collection tube, counting, based on the image data and using a model trained to distinguish between an upright state of the blood collection tube and an upside down state of the blood collection tube, a number of inversions performed on the blood collection tube; and 
 in response to determining the number of inversions performed is less than a required number of inversions, issuing an alert indicating that insufficient inversions were performed. 
 
     
     
       14. The non-transitory machine-readable medium of  claim 13 , wherein identifying whether the blood collection tube is present in the image data includes comparing a feature identified in the image data to features of blood collection tubes in a memory. 
     
     
       15. The non-transitory machine-readable medium of  claim 13 , wherein identifying whether the blood collection tube includes blood therein includes determining whether a body portion of the blood collection tube has changed color in the image data. 
     
     
       16. The non-transitory machine-readable medium of  claim 13 , wherein the operations further comprise identifying a number of inversions required to be performed on the blood collection container. 
     
     
       17. The non-transitory machine-readable medium of  claim 16 , wherein identifying the number of inversions required includes identifying a color of a cap of the blood collection container and looking up the number of inversions associated with the identified color in the memory. 
     
     
       18. The non-transitory machine-readable medium of  claim 17 , wherein identifying the number of inversions required includes identifying whether a material is present in a body portion of the blood collection container. 
     
     
       19. A method for phlebotomist monitoring, the method comprising:
 identifying, by object recognition implemented by processing circuitry, whether a blood collection tube is present in image data from a camera situated to capture images of a phlebotomist collecting a sample; 
 in response to identifying the blood collection tube is present in the field of view of the camera based on the image data, identifying, by the processing circuitry, whether the blood collection tube includes blood therein; 
 after identifying the blood is present in the blood collection tube, counting, based on the image data and by a computer model trained to identify when the blood collection tube is in an upright state and when the blood collection tube is in an upside down state, a number of inversions performed on the blood collection tube; and 
 in response to determining the number of inversions performed is less than a required number of inversions, issuing, by communications circuitry, an alert indicating that insufficient inversions were performed. 
 
     
     
       20. The method of  claim 19 , wherein counting the number of inversions performed on the blood collection tube includes identifying a first feature of the blood collection tube on a cap of the blood collection tube and a second feature on an end portion of the blood collection tube opposing the cap, and for each inversion, determining that pixels corresponding to the first feature are vertically displaced from the second feature in a first direction and then vertically displace from the second feature in a second, opposite direction in image data corresponding to a subsequent image.